DESCRIPTION (Applicant's Description) We propose to conduct an exploratory investigation of the Web's content on cancer. We ask one overarching, and deceptively simple, research question: For information on cancer, what are the attributes of the Web as an information source? What's out there? To answer this question, we will address issues of structure and content. To address structure, we will make use of emerging technology in web search engines to develop a taxonomy of the cancer information resources available on the web. This work will be done in collaboration with researchers in the IBM Almaden Research Center (the Clever Project). This taxonomic analysis will be designed to reveal the large-scale structure of cancer information on the Web, its interlinkages, and its natural "information communities." This analysis will also be executed at two points in time (about a year apart) so that an evaluation of change may be made. To address content, we will make use of the results of the Clever search (with appropriate sampling and sub-grouping strategies) to execute four studies. First, we will use a variety of published criteria for the evaluation of health information Web sites to describe the overall characteristics of this public resource (e.g., prevalence of the use of visuals, interactivity, linkages, source attribution, availability to non-English readers, etc.). Second, we will execute a simple study of the readability of the text in this domain (using standard text assessment tools such as FOG and SMOG). Third, we will conduct a study to carefully examine the visual content of this resource. And fourth, we will examine a sample of prevention-oriented Web sites to specifically assess how cancer risk information is framed (with emphasis on nutrition, tobacco use, and sun exposure). The outcome of this research will be an intensive look at a widely available health information resource, the characteristics of which are not empirically well known at present. This study will add to the small but growing body of research focusing on the quality of cancer information on the Web, will serve as a baseline for those wishing to monitor cancer information on the Web, and will be a useful guide for both users and developers of these Web resources. Further, this study will serve as a foundation for our (and others') intended future work examining the quality of cancer information on the Web -- and the experiences had by users of this public resource. It is also our hope--and intention--that the results of this study will also be useful to physicians who are grappling with questions involving how to direct patients toward useful and accurate information about cancer prevention and treatment in a rapidly evolving on-line environment. And we hope that any improvements that our study may foster in the form and content of cancer information on the Web will provide benefits to the most important users of this public resource: the patients.